Modeling and prediction of regional sound speed profile grid using eXtreme Gradient Boosting model
Physica Scripta · 26.06.2026 · 101 (25) · 256007
DOI: 10.1088/1402-4896/ae783e ↗Kurz gefasst
Fachartikel in der Zeitschrift Physica Scripta, online erschienen am 24.06.2026. Verfasst von Zhen Li und 3 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Sound speed, as a fundamental parameter characterizing ocean acoustics, is the prerequisite for advancing the research and application of underwater acoustic technologies. This study examines the use of a machine learning model to reflect the spatial and temporal correlation between the points in a regional sound speed profile (SSP) grid and to make SSP predictions. The input and output data to the machine learning algorithm consist of SSP data. For the prediction experiments, the eXtreme Gradient Boosting (XGBoost) algorithm is introduced and evaluated using the SSP dataset collected from 25 points. The results indicate that the prediction accuracy of XGBoost has increased by 21.43% compared with the decision tree model, and the root mean square error of the sound speed time series and SSP prediction results are 0.41 m s –1 and 0.45 m s –1 , respectively. The experiment with a spatial resolution of 2.0° × 2.0° demonstrates that this study framework can improve the spatial resolution of SSP data with fewer observations. Overall, the XGBoost model demonstrates excellent performance in SSP prediction and can provide data support for quality correction of SSP and related research.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Li, Z., Chai, H., Lu, T., & Li, N. (2026). Modeling and prediction of regional sound speed profile grid using eXtreme Gradient Boosting model. Physica Scripta, 101(25), 256007. https://doi.org/10.1088/1402-4896/ae783e
Vancouver
Li Z, Chai H, Lu T, Li N. Modeling and prediction of regional sound speed profile grid using eXtreme Gradient Boosting model. Physica Scripta. 2026;101(25):256007. doi:10.1088/1402-4896/ae783e
BibTeX
@article{Li2026,
author = {Li, Zhen and Chai, Hongzhou and Lu, Tieding and Li, Na},
title = {Modeling and prediction of regional sound speed profile grid using eXtreme Gradient Boosting model},
journal = {Physica Scripta},
year = {2026},
volume = {101},
number = {25},
pages = {256007},
doi = {10.1088/1402-4896/ae783e}
}
EUVIMED